Sources of the Implicit Priming Effect of the Syllable in Chinese Word Production

Jenn-Yeu Chen and Train-Min Chen
National Chung-Cheng University, Taiwan

The implicit priming task has been a useful tool for studying word-form encoding in word production. In the task, the participants first learn a set of word pairs. Then, the first word (prompt) of each pair is presented one at a time, and the participants have to say the second word (target) immediately. The target words are manipulated in such a way that in one condition, they form several homogeneous blocks, wherein the words share a particular morphophonological property (i.e., the implicit prime, e.g., the first syllable), and in the other condition, the same words form the heterogeneous blocks, wherein the words do not share the specified property. The response times are typically faster in the homogeneous blocks than in the heterogeneous blocks. The facilitation, commonly referred to as the implicit priming (IP) effect, reflects the preparation of some process(es) in word production, but the source(s) of the effect has remained unclear. Meyer (1990) suggested that the effect reflected the priming of the segments and the segment-to-frame association. Levelt, Meyer, and Roelofs (1999) speculated that the effect had to do with retrieval from the mental syllabary. We conducted several implicit priming experiments with the syllable alone (8 experiments with 28 syllables) or the syllable plus the tone (6 experiments with 23 syllables) as the implicit primes. The syllables contained from one to three segments. The IP effect was, on the average, 10 ms for the syllable alone prime, and 45 ms for the syllable+tone prime. A meta-analysis was performed to examine how the IP effects varied as a function of the type of the prime and the length of the syllable. The results showed two regression lines with similar slopes (8.32 and 10.68) but different intercepts (-9.56 and 22.18) for the syllable alone and the syllable+tone primes, respectively. Further analyses included the frequencies of the syllables. The results showed that syllable length was the only contributing factor in the syllable alone IP effect, but both syllable length and syllable+tone frequency contributed to the syllable+tone IP effect. These findings support at least two distinct processes in word-form encoding, one being the segment-to-frame association and the other the retrieval of the phonetic syllabary. The IP effect of the syllable alone is the result of preparing the first process. The IP effect of the syllable plus the tone is the combined result of preparing both processes.